Occupational exposure to manganese has been in the news lately, with law suits by welders who claim neurological disease caused by manganese exposure. Now two scientists at Sweden’s Karolinska Institute have written a paper in which they argue that current guidelines for safe levels of manganese in drinking water are based on a misinterpretation of a twenty-five year old study, and that newer evidence suggests that at least for infants and other vulnerable populations, the current guideline values are not adequately protective.

In a paper available online at Environmental Health Perspectives, Karin Ljung and Marie Vahter trace back the foundation for the World Health Organization’s (and the EPA’s) recommendation for manganese in drinking water to a single study from 1982 that was misinterpreted in calculating a No Observed Adverse Effect Level. That mistake, combined with several new studies showing neurological effects in children, lead the authors to conclude that it’s time to re-evaluate the guideline data.

Manganese is a trace element that occurs naturally in soil, water, and plants, and we need a small amount of it in our diet – but not too much. Several studies have linked excessive manganese exposure and neurological disorders in children. Infants are of particular concern because they retain more manganese than adults do, probably because they absorb more and excrete less of it. Breast-fed infants tend to have a low exposure, but infant formulas contain manganese concentrations around 100-fold higher than those of breast milk. If powdered infant formulas are mixed with drinking water containing manganese equivalent to the WHO guideline value, the limit set for infant formula by the European Commission’s Scientific Committee on Food can be exceeded.

In 2006, the WHO lowered the guideline value for manganese in drinking water from 500 to 400 μg/L. Ljung and Vahter explain how WHO decided on that number:

The current health-based guideline value of 400 μg L for manganese in water is based on a estimated No Observed Adverse Effect Level (NOAEL) for manganese in food. In order to allow for the possible higher bioavailability of manganese when ingested with water than with food (WHO 2004), the NOAEL of 11 mg/day was divided by an uncertainty factor of three. Using an adult body weight of 60 kg, a tolerable daily intake (TDI) of 60 μg manganese per kg bodyweight was derived. On the assumption that 20% of the TDI would be allowed to come from drinking-water, and that an adult consumes two liters of water per day, the value of 400 μg/L was set as a health-based guideline value for drinking water (WHO 2004).

WHO cited a 1999 review by J.L. Greger – and an Institute of Medicine Report that references Greger – that indicated that no adverse effects were observed in people whose diets could be giving them manganese intakes as high as 10.9 mg/day. The Greger article is not the best source for setting a guideline value for drinking water, Ljung and Vahter suggest (emphasis added):

However, the article by Greger (1999) did not focus on manganese intake from different diets but rather on potential biomarkers of manganese in human nutrition and toxicology. In her introduction section on manganese exposure she stated that “the average intakes of adults eating Western-type and vegetarian diets in various surveys ranged from 0.7-10.9 mg Mn/day” with reference to Gibson (1994) and Freeland-Graves (1994). However, no mentioning of these values is made in Gibson (1994), whose article is a review on trace elements in vegetarian and omnivorous diets with a concern for deficiencies in vegetarian diets. The Freeland-Graves (1994) article is also a review paper published in a book on risk assessment of essential elements by the International Life Science Institute. The intake values of 0.7-10.8 mg Mn/day were observed for Canadian women in a study carried out by Gibson and Scythes in 1982 and were in the Freeland-Graves (1994) review presented in a table of daily manganese intakes compiled from several studies.

The values used for setting the NOAEL thus originate from one study, where one hundred Canadian women aged 30 ± 6.1 years were asked to complete dietary protocols of all consumed foods and beverages (including drinking water) in their own homes for three consecutive weekdays. The authors present both calculated and analysed intake values. The calculated daily manganese intake ranged from 0.7 to 10.8 mg, where 90% of the women ingested less than 5 mg/day, and almost half of the women (40%) ingested less than 2.5 mg manganese per day. The average daily manganese intake was calculated at 3.1 ± 1.5 mg. The analysed manganese intake from duplicate portions provided a slightly lower average daily manganese intake of 2.4 mg/day. The analysed maximum intake was not presented. The slight discrepancy between the calculated and the analysed manganese intakes were explained by the authors to be a result of an over-estimation in portion size (Gibson and Scythes 1982).

The NOAEL of 11 mg/day is thus based on calculated daily intakes of manganese, and not on actual measurements of manganese intakes. No mention is made of the subjects’ health status, why it seems unfounded to draw any conclusions on a “no observed adverse effect level” of daily manganese intakes at 11 mg. In fact, several studies from different countries have reported daily manganese intakes after 1982, when the Gibson and Scythes study was published. … None of the studies found intakes as high as 11 mg/day.

Ljung and Vahter also question some of the calculations behind the Lowest Observed Adverse Effect Level (LOAEL) for manganese in water. They conclude that:

the maximum content of manganese in infant formula is not based on scientific data on infants, but is instead based on studies of adults, which also seem to have been misinterpreted. It should also be noted that the maximum daily manganese intake from formula (500 μg/day) is about two orders of magnitude higher than the calculated adequate daily intake of 3 μg manganese for infants up to 6 months, which is based on the concentration in breast milk. The bioavailability of manganese in drinking water and formula is lower than that in breast milk, which may justify a higher concentration. However, it is questionable whether a 100-fold higher manganese concentration in formula is justifiable.

Ljung and Vahter make a strong case for re-evaluating the data used in calculating the guideline for manganese in drinking water. A lower value may not be required for the majority of the population, but reading this paper, is does seem that that sensitive groups, including infants, may be at risk with the present guideline value.